Control Channel Isolation in SDN Virtualization: A Machine Learning Approach

Yeonho Yoo; Gyeongsik Yang; Changyong Shin; Jeunghwan Lee; Chuck Yoo

doi:10.1109/CCGrid57682.2023.00034

Control Channel Isolation in SDN Virtualization: A Machine Learning Approach

Yeonho Yoo, Gyeongsik Yang, Changyong Shin, Jeunghwan Lee, Chuck Yoo

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

5 Citations (Scopus)

Abstract

Performance isolation is an essential property that network virtualization must provide for clouds. This study addresses the performance isolation of the control plane in virtualized software-defined networking (SDN), which we call control channel isolation. First, we report that the control channel isolation is seriously broken in the existing network hypervisor in that the end-to-end control latency grows by up to 15 x as the number of virtual switches increases. This jeopardizes the key network operations, such as routing, in datacenters. To address this issue, we take a machine learning approach that learns from the past control traffic as time-series data. We propose a new network hypervisor, Meteor, that designs an LSTM autoencoder to predict the control traffic per virtual switch. Our evaluation results show that Meteor improves the processing latency per control message by up to 12.7x. Furthermore, Meteor reduces the end-to-end control latency by up to 73.7%, which makes it comparable to the non-virtualized SDN.

Original language	English
Title of host publication	Proceedings - 23rd IEEE/ACM International Symposium on Cluster, Cloud and Internet Computing, CCGrid 2023
Editors	Yogesh Simmhan, Ilkay Altintas, Ana-Lucia Varbanescu, Pavan Balaji, Abhinandan S. Prasad, Lorenzo Carnevale
Publisher	Institute of Electrical and Electronics Engineers Inc.
Pages	273-285
Number of pages	13
ISBN (Electronic)	9798350301199
DOIs	https://doi.org/10.1109/CCGrid57682.2023.00034
Publication status	Published - 2023
Event	23rd IEEE/ACM International Symposium on Cluster, Cloud and Internet Computing, CCGrid 2023 - Bangalore, India Duration: 2023 May 1 → 2023 May 4

Publication series

Name	Proceedings - 23rd IEEE/ACM International Symposium on Cluster, Cloud and Internet Computing, CCGrid 2023

Conference

Conference	23rd IEEE/ACM International Symposium on Cluster, Cloud and Internet Computing, CCGrid 2023
Country/Territory	India
City	Bangalore
Period	23/5/1 → 23/5/4

Bibliographical note

Publisher Copyright:
© 2023 IEEE.

Keywords

Control channel
Isolation
LSTM autoencoder
Machine learning
Network virtualization
SDN

ASJC Scopus subject areas

Artificial Intelligence
Computer Networks and Communications
Control and Optimization
Hardware and Architecture
Information Systems

Access to Document

10.1109/CCGrid57682.2023.00034

Cite this

Yoo, Y., Yang, G., Shin, C., Lee, J., & Yoo, C. (2023). Control Channel Isolation in SDN Virtualization: A Machine Learning Approach. In Y. Simmhan, I. Altintas, A.-L. Varbanescu, P. Balaji, A. S. Prasad, & L. Carnevale (Eds.), Proceedings - 23rd IEEE/ACM International Symposium on Cluster, Cloud and Internet Computing, CCGrid 2023 (pp. 273-285). (Proceedings - 23rd IEEE/ACM International Symposium on Cluster, Cloud and Internet Computing, CCGrid 2023). Institute of Electrical and Electronics Engineers Inc.. https://doi.org/10.1109/CCGrid57682.2023.00034

Control Channel Isolation in SDN Virtualization: A Machine Learning Approach. / Yoo, Yeonho; Yang, Gyeongsik; Shin, Changyong et al.
Proceedings - 23rd IEEE/ACM International Symposium on Cluster, Cloud and Internet Computing, CCGrid 2023. ed. / Yogesh Simmhan; Ilkay Altintas; Ana-Lucia Varbanescu; Pavan Balaji; Abhinandan S. Prasad; Lorenzo Carnevale. Institute of Electrical and Electronics Engineers Inc., 2023. p. 273-285 (Proceedings - 23rd IEEE/ACM International Symposium on Cluster, Cloud and Internet Computing, CCGrid 2023).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

Yoo, Y, Yang, G, Shin, C, Lee, J & Yoo, C 2023, Control Channel Isolation in SDN Virtualization: A Machine Learning Approach. in Y Simmhan, I Altintas, A-L Varbanescu, P Balaji, AS Prasad & L Carnevale (eds), Proceedings - 23rd IEEE/ACM International Symposium on Cluster, Cloud and Internet Computing, CCGrid 2023. Proceedings - 23rd IEEE/ACM International Symposium on Cluster, Cloud and Internet Computing, CCGrid 2023, Institute of Electrical and Electronics Engineers Inc., pp. 273-285, 23rd IEEE/ACM International Symposium on Cluster, Cloud and Internet Computing, CCGrid 2023, Bangalore, India, 23/5/1. https://doi.org/10.1109/CCGrid57682.2023.00034

Yoo Y, Yang G, Shin C, Lee J, Yoo C. Control Channel Isolation in SDN Virtualization: A Machine Learning Approach. In Simmhan Y, Altintas I, Varbanescu AL, Balaji P, Prasad AS, Carnevale L, editors, Proceedings - 23rd IEEE/ACM International Symposium on Cluster, Cloud and Internet Computing, CCGrid 2023. Institute of Electrical and Electronics Engineers Inc. 2023. p. 273-285. (Proceedings - 23rd IEEE/ACM International Symposium on Cluster, Cloud and Internet Computing, CCGrid 2023). doi: 10.1109/CCGrid57682.2023.00034

Yoo, Yeonho ; Yang, Gyeongsik ; Shin, Changyong et al. / Control Channel Isolation in SDN Virtualization : A Machine Learning Approach. Proceedings - 23rd IEEE/ACM International Symposium on Cluster, Cloud and Internet Computing, CCGrid 2023. editor / Yogesh Simmhan ; Ilkay Altintas ; Ana-Lucia Varbanescu ; Pavan Balaji ; Abhinandan S. Prasad ; Lorenzo Carnevale. Institute of Electrical and Electronics Engineers Inc., 2023. pp. 273-285 (Proceedings - 23rd IEEE/ACM International Symposium on Cluster, Cloud and Internet Computing, CCGrid 2023).

@inproceedings{8223c8656a1e4478bbdce7cc08aaec68,

title = "Control Channel Isolation in SDN Virtualization: A Machine Learning Approach",

abstract = "Performance isolation is an essential property that network virtualization must provide for clouds. This study addresses the performance isolation of the control plane in virtualized software-defined networking (SDN), which we call control channel isolation. First, we report that the control channel isolation is seriously broken in the existing network hypervisor in that the end-to-end control latency grows by up to 15 x as the number of virtual switches increases. This jeopardizes the key network operations, such as routing, in datacenters. To address this issue, we take a machine learning approach that learns from the past control traffic as time-series data. We propose a new network hypervisor, Meteor, that designs an LSTM autoencoder to predict the control traffic per virtual switch. Our evaluation results show that Meteor improves the processing latency per control message by up to 12.7x. Furthermore, Meteor reduces the end-to-end control latency by up to 73.7%, which makes it comparable to the non-virtualized SDN.",

keywords = "Control channel, Isolation, LSTM autoencoder, Machine learning, Network virtualization, SDN",

author = "Yeonho Yoo and Gyeongsik Yang and Changyong Shin and Jeunghwan Lee and Chuck Yoo",

note = "Publisher Copyright: {\textcopyright} 2023 IEEE.; 23rd IEEE/ACM International Symposium on Cluster, Cloud and Internet Computing, CCGrid 2023 ; Conference date: 01-05-2023 Through 04-05-2023",

year = "2023",

doi = "10.1109/CCGrid57682.2023.00034",

language = "English",

series = "Proceedings - 23rd IEEE/ACM International Symposium on Cluster, Cloud and Internet Computing, CCGrid 2023",

publisher = "Institute of Electrical and Electronics Engineers Inc.",

pages = "273--285",

editor = "Yogesh Simmhan and Ilkay Altintas and Ana-Lucia Varbanescu and Pavan Balaji and Prasad, {Abhinandan S.} and Lorenzo Carnevale",

booktitle = "Proceedings - 23rd IEEE/ACM International Symposium on Cluster, Cloud and Internet Computing, CCGrid 2023",

}

TY - GEN

T1 - Control Channel Isolation in SDN Virtualization

T2 - 23rd IEEE/ACM International Symposium on Cluster, Cloud and Internet Computing, CCGrid 2023

AU - Yoo, Yeonho

AU - Yang, Gyeongsik

AU - Shin, Changyong

AU - Lee, Jeunghwan

AU - Yoo, Chuck

PY - 2023

Y1 - 2023

N2 - Performance isolation is an essential property that network virtualization must provide for clouds. This study addresses the performance isolation of the control plane in virtualized software-defined networking (SDN), which we call control channel isolation. First, we report that the control channel isolation is seriously broken in the existing network hypervisor in that the end-to-end control latency grows by up to 15 x as the number of virtual switches increases. This jeopardizes the key network operations, such as routing, in datacenters. To address this issue, we take a machine learning approach that learns from the past control traffic as time-series data. We propose a new network hypervisor, Meteor, that designs an LSTM autoencoder to predict the control traffic per virtual switch. Our evaluation results show that Meteor improves the processing latency per control message by up to 12.7x. Furthermore, Meteor reduces the end-to-end control latency by up to 73.7%, which makes it comparable to the non-virtualized SDN.

AB - Performance isolation is an essential property that network virtualization must provide for clouds. This study addresses the performance isolation of the control plane in virtualized software-defined networking (SDN), which we call control channel isolation. First, we report that the control channel isolation is seriously broken in the existing network hypervisor in that the end-to-end control latency grows by up to 15 x as the number of virtual switches increases. This jeopardizes the key network operations, such as routing, in datacenters. To address this issue, we take a machine learning approach that learns from the past control traffic as time-series data. We propose a new network hypervisor, Meteor, that designs an LSTM autoencoder to predict the control traffic per virtual switch. Our evaluation results show that Meteor improves the processing latency per control message by up to 12.7x. Furthermore, Meteor reduces the end-to-end control latency by up to 73.7%, which makes it comparable to the non-virtualized SDN.

KW - Control channel

KW - Isolation

KW - LSTM autoencoder

KW - Machine learning

KW - Network virtualization

KW - SDN

UR - http://www.scopus.com/inward/record.url?scp=85166323359&partnerID=8YFLogxK

U2 - 10.1109/CCGrid57682.2023.00034

DO - 10.1109/CCGrid57682.2023.00034

M3 - Conference contribution

AN - SCOPUS:85166323359

T3 - Proceedings - 23rd IEEE/ACM International Symposium on Cluster, Cloud and Internet Computing, CCGrid 2023

SP - 273

EP - 285

BT - Proceedings - 23rd IEEE/ACM International Symposium on Cluster, Cloud and Internet Computing, CCGrid 2023

A2 - Simmhan, Yogesh

A2 - Altintas, Ilkay

A2 - Varbanescu, Ana-Lucia

A2 - Balaji, Pavan

A2 - Prasad, Abhinandan S.

A2 - Carnevale, Lorenzo

PB - Institute of Electrical and Electronics Engineers Inc.

Y2 - 1 May 2023 through 4 May 2023

ER -

Control Channel Isolation in SDN Virtualization: A Machine Learning Approach

Abstract

Publication series

Conference

Bibliographical note

Keywords

ASJC Scopus subject areas

Access to Document

Other files and links

Fingerprint

Cite this